Vital processes in living organisms are closely associated with reactive species, which represent a very diverse group from the point of view of both their chemistry and biological impact. These include molecules, atoms, and ions. Some of them belong to a family known as highly reactive free radicals. In aerobic organisms, many reactive species derive from molecular oxygen, some of them are nitrogen- and sulfur-containing, or have other chemical features, for example reactive carbonyl group(s). Although reactive species can be produced endogenously, for instance, as side products of enzymatic metabolism, they are often involved in nonenzymatic reactions (1–4). Unlike enzymatic reactions, which are precisely controlled by respective enzymes, nonenzymatic processes are rather unspecific and occur spontaneously. Generally, nonenzymatic events and reactive species are well-known for playing a detrimental role associated with health risks. On the other hand, being an important part of immune response, regulators of gene expression, and cellular signaling messengers, reactive species at low concentrations can have beneficial effects (5–9).

Historically, most studies in the field of reactive species and nonenzymatic processes in living organisms were focused on free radical oxidation and glycation, related to reactive oxygen species (ROS) and reactive carbonyl species (RCS) in particular. Increased levels of ROS and RCS lead to development of oxidative and carbonyl stresses, respectively. Obviously, both nonenzymatic processes mentioned above are closely linked to each other—oxidation reactions and ROS are found to be involved in certain stages of glycation, at the same time glycation itself is a source of ROS and ROS-mediated reactions (10–12). To reflect the combinational impact of nonenzymatic oxidation and glycation, the term “glycoxidation” has been introduced and is commonly used (13, 14). In addition, another nonenzymatic process such as ROS-induced lipid peroxidation results in a variety of RCS and respective glycation products, demonstrating an interplay between oxidative and carbonyl stresses. Glycoxidation leads to irreversible modifications of biomolecules and damages to cellular constituents with the consequent formation of poorly degraded compounds, collectively named advanced glycation end products (AGEs). Elevated levels of AGEs initiate activation of the specific receptor for AGEs (RAGE), mediating and amplifying inflammatory events related to metabolic disorders, chronic diseases, and aging (4, 15, 16). In this regard, inhibiting the AGE-RAGE axis is considered as a logical target in pathological conditions.

Reducing carbohydrates are major contributors to glycation process under physiological conditions. Among other monosaccharides, fructose has been found to be one of the most potent glycation agents in vivo (17–20). Numerous studies have demonstrated adverse metabolic changes in humans and animals as deleterious side effects of long-term intake of excessive fructose. These effects could be attributed to fructose-mediated glycation, increased levels of RCS and ROS, AGE accumulation, and subsequent activation of RAGE (18, 21–26).

Importantly, there is compelling evidence that short-term application of moderate doses of fructose can have beneficial defensive effects under certain pathophysiological conditions, particularly those associated with oxidative/carbonyl stress (21, 23, 27–30). Therefore, strategies such as diet optimization—limiting added sugars, and fructose in particular—could effectively prevent overactivation of RAGE signaling, and thereby mitigate the detrimental impacts of fructose. Other mild approaches may also offer moderate modulation of the fructose-mediated AGE-RAGE axis.

The initial step of fructose-initiated glycation (Figure 1) is a covalent interaction between electrophilic carbonyl group of open-chain fructose and nucleophilic functional group of a biomolecule. This interaction produces a diverse group of early glycation products (Schiff bases and Heyns products). Schiff bases are relatively unstable compounds that can be subjected to further isomerization (Heyns rearrangement) and formation of more stable Heyns adducts. Amadori products derived from glucose are also belong to early glycation products. The fructose moiety of the Heyns products can undergo enolization followed by dehydration, oxidation, and/or fragmentation reactions, consequently producing a variety of reactive species (10, 13, 31).

Besides glycation, other nonenzymatic reactions contribute to the formation of reactive species. They are autoxidation of glucose, fructose and other monosaccharides (Wolff pathway), autoxidation of Amadori and Heyns compounds (Hodge pathway), and oxidative fragmentation of Schiff bases (Namiki pathway) (2, 11, 14, 32, 33). Some fructose-derived RCS have been identified in vivo (Figure 2), for example, glycolaldehyde and acrolein. Importantly, most biological damages caused by RCS are related to α-dicarbonyls (compounds with two adjacent reactive carbonyl groups), including glyoxal, methylglyoxal, 3-deoxyglucosone, 3-deoxyfructose. This is likely because these compounds are about 20,000-fold more reactive than fructose and other reducing carbohydrates (31, 34, 35).

In the advanced stages of glycation, early products of the process and reactive species further undergo complex of parallel, sequential, and branched chains of irreversible nonenzymatic reactions, ultimately producing a variety of AGEs. Fructose-derived AGEs have been found in various tissues and peripheral blood. Common AGEs detected in vivo include pentosidine, glucosepane, argpyrimidine, methylglyoxal-lysine dimmer, glyoxal-lysine dimmer, carboxymethylcysteine, carboxymethylguanosine, carboxymethyllysine, glycolaldehyde-pyridine, 3-deoxyglucosone-lysine dimer, imidizalone, pyrraline, etc. (23, 25, 26). The chemical structures of some AGEs are presented in Figure 3. Generally, AGEs are highly heterogeneous group of compounds with different structures, sizes, and molecular weights. They exhibit various physico-chemical properties; for example, some of AGEs, like pentosidine, are fluorescent, while others, such as methylglyoxal- and glyoxal-lysine dimmers, are nonfluorescent. Also, AGEs may demonstrate different cross-linking abilities (14, 36).

Formation of glycation products is a relatively slow process. In contrast to rapid generation of products in free-radical chain oxidation, AGE formation takes weeks, months, or even years. This could be explained by differences in average reactivities of RCS and ROS and their half-life times. While half-lives of many RCS range from minutes to hours, the most reactive ROS have half-life periods from 10⁻⁹ to 10⁻⁶ s (37–39). The slow and irreversible modification of biomolecules via glycation correlates with aging. A causative role of glycation in the aging process and age-associated pathologies has been suggested (40). As confirmation, AGE-modified albumin has been reported to be specifically recognized and degraded by macrophages in vitro. Therefore, AGEs have been supposed to serve as specific signals for the recognition and degradation of senescent macromolecules in vivo. On the other hand, incomplete removal of glycated proteins by macrophages was connected with aging (41). In fact, the free-radical theory of aging by Harman (42) and the glycation hypothesis of aging by Monnier and Cerami (40) are currently the most popular and widely accepted basis to explain molecular mechanisms of aging.

During glycation, the appearance of reactive species is an unavoidable step toward the formation of AGEs. In addition to this “main direction” (Figure 1, blue arrows), reactive carbonyls can return to the starting point and initiate new rounds of glycation, since RCS have at least one carbonyl group to interact with new biomolecules. It should be noted that most biomolecules are nucleophiles, as many of them, including proteins, amino lipids, nucleic acids, possess amino, sulfhydryl, guanidine or other nucleophilic functional groups. Thus, many different biomolecules are involved in the irreversible formation of AGEs. The latters, in turn generate both RCS and ROS (33), leading to new stages of glycoxidation, creating more and more vicious nonenzymatic chemical cycles (Figure 1, red arrows).

In addition to nonenzymatic processes, normal enzymatic metabolism can also produce reactive species, including RCS. Under physiological conditions, up to 0.4% of glycolytic intermediates can be converted to methylglyoxal, a side product of glycolysis (43). Under other circumstances, reactive carbonyls are generated enzymatically as necessary products. In activated neutrophils, for example, some amino acids are found to be oxidized to different RCS, that demonstrates particular importance of reactive carbonyls in the immune response (44–46). Also, enzymatically produced RCS are involved in regulating gene expression and signaling networks (5, 7, 39). It is important to note that reactive species play a beneficial role only at their low steady-state concentrations, which are tightly controlled by appropriate enzymes in the organism.

Nonetheless, some enzymatic pathways can intensify RCS production to relatively high levels, when reactive carbonyls demonstrate detrimental effects. For example, the polyol pathway, which is responsible for converting glucose to fructose, increases under hyperglycemia in the insulin-insensitive tissues. As mentioned, fructose itself may initiate glycation and be converted to RCS. In both humans and animal models, increased fructose and enhanced flux of polyols have been suggested as metabolic basis for different pathologies (47, 48).

Importantly, high fructose diet, as a source of exogenous fructose, is clearly related to deleterious side effects, including metabolic syndrome, kidney and heart diseases, insulin resistance and type 2 diabetes, nonalcoholic fatty liver disease, arthritis, dementia and neurodegenerative disorders. Glycation is believed to contribute significantly to the fructose-mediated pathogenesis. As confirmation, chronic exposure to high exogenous fructose has been observed to stimulate RCS/ROS formation and AGE accumulation in various in vitro and in vivo experimental models (17, 22, 32, 49–51). Moreover, excessive fructose and glycation products can alter gut microbiota composition, potentially leading to inflammatory events. Overconsumption of fructose (above 50 g per day) is one of the major factors contributing to chronic inflammation and related diseases (20, 52).

Conversely, epidemiological, clinical, and experimental studies suggest that fructose intake of less than 50 g per day is a moderate dose that may have no detrimental influence (53–56). Moreover, clinical investigations evidenced that small doses of fructose (≤10 g/meal) significantly decreased the postprandial glycemic response to high glycemic index foods and demonstrated beneficial effect on the levels of HbA1c and triglycerides, as well as improved blood pressure and body mass index (53, 57–60). In addition, the Dietary Guidelines for Americans recommend limiting calories from added sugars to less than 10% of total daily calories (61). This recommendation corresponds well to the mentioned above moderate doses of fructose.

According to Marriott and colleagues (62), the highest mean percentage of added fructose intake is associated to the consumption of nonalcoholic beverages (54.3%), grain products (20.3%), and sweets (12.1%). Also, the highest mean percentage of total fructose consumption is linked to nonalcoholic beverages (46.0%), grain products (17.3%), fruits and fruit products (13.4%), sweets (10.3%), milk and milk products (7.1%), and vegetables and vegetable products (2.7%).

Most foods contain reactive carbonyls and AGEs, in fact, in substantive amounts as compared to their endogenous sources, making them a major exogenous source of glycation products, known as dietary glycotoxins (61–67). For example, the amount of ingested Heyns products (calculated as fructoselysine) per day ranges between 500 and 1,200 mg, while the quantity of dietary AGEs (mainly pyrraline and carboxymethyllysine) ranges from 25 to 75 mg (68). High-fructose corn syrup, a widely used sweetener in many foods and beverages, is the major source of such RCS as 3-deoxyglucosone, glyoxal, and methylglyoxal (69–71). Consequently, the Western diet, which is rich in processed and sweetened foods, is strongly associated with a high intake of dietary glycotoxins.

Several comprehensive works demonstrate the content of AGEs in commonly consumed foods, providing a solid foundation for creating dietary AGE database (67, 72–74). Unfortunately, there are no standardized conventional methods for measuring dietary AGEs and other glycotoxins, which limits inter-laboratory comparisons of the data obtained. Nevertheless, the AGE levels in common foods are examined and some studies report how AGE content varies by food category and cooking method. It has been shown that total amounts of AGEs were high in processed nuts, bakery products, and certain types of meats and cereals (>150 mg/kg), while AGE content is rather low in dairy products, vegetables, and fruits (< 40 mg/kg) (73).

In fact, AGEs are unavoidable components of our daily meals, as they are present in natural, uncooked foods. However, processed foods—especially those high in fat, sugar, and protein (Western diet)—tend to be particularly rich in AGEs. Any food processing increases AGE content in regime-dependent manner during cooking. It has been well demonstrated that frying, broiling, grilling, and roasting substantially accelerated dietary AGE formation, whereas foods cooked by steaming, stewing, poaching, or boiling had relatively low amounts of AGEs. For example, grilled, boiled, and raw carrots had approximately 220, 25, and 10 AGE kU/100 g, respectively. At the same time, fried, poached, and raw chicken contained about 7,000, 1,000, and 800 AGE kU/100 g (67).

Consumption of AGE-rich diet, in which the amounts of AGEs were 3-fold higher than in a regular diet, has been shown to elevate blood concentration of AGEs by about 1.5-fold (75). It has been also observed that up to 30% of dietary AGEs were absorbed in human organism, two-thirds of the absorbed AGEs remained in the body and one-third was excreted during the next 3 days (75–77). Interestingly, the comparison of AGE concentrations in plasma of vegetarian and omnivorous individuals presented significantly higher (by 25%) AGE levels in vegetarians. In comparison with the omnivores, vegans, lacto-ovo-vegetarians, and semi-vegetarians demonstrated the levels of some AGEs higher by 15%, 32%, and 22%, respectively (18, 78).

As noted, some biological effects of AGEs result from their involvement in generating RCS/ROS (Figure 1). An increase in the levels of reactive species leads, in turn, to chronic carbonyl/oxidative stress development and modification of biomolecules with consequent loss of their functions. Also, ingested AGEs elicit toxicological effects by disrupting gut microbiome and immune homeostasis (79, 80). Both exogenously and endogenously derived AGEs are responsible for modulating certain signaling pathways. In particular, they bind to RAGE (Figure 1) and activate the RAGE-downstream pathway, which is involved in a variety of biological processes from proliferation and inflammation to apoptosis and autophagy (15, 20, 81). Upregulation of RAGE is known mostly for playing a detrimental role associated with various pathologies.

The Receptor for AGEs (RAGE) is a member of the immunoglobulin superfamily of cell-surface molecules. Originally purified from bovine lung endothelial cells (82, 83), RAGE is now known to be an ubiquitous receptor present in a variety of both immune and nonimmune cells (leucocytes, monocytes, macrophages, pericytes, and endothelial cells, etc.). RAGE is found in various organs and tissues such as the heart, liver, lungs, kidneys, skeletal muscles, and, the cardiovascular and nervous systems (84). Besides AGEs, RAGE recognizes a number of other ligands: (1) damage-associated molecular pattern molecules (DAMPs)-amyloid-β peptides (Aβ), β-sheet fibrils, high mobility group box 1 (HMGB1), and S100/calgranulinprotein; (2) pathogen-associated molecular pattern (PAMP) lipopolysaccharide (LPS); and (3) DAMP/PAMP DNA and RNA species (15, 85, 86). The above demonstrates that RAGE is the multiligand receptor, which is able to bind various compounds with different structures, sizes, functions, stimulating thereby multiple downstream signaling cascades.

What molecular mechanisms enable RAGE to recognize such a wide range of ligands and contribute to its diverse bioeffects? Structural studies reveal that full-length human RAGE has three domains (Figure 4): an extracellular part responsible for ligand binding, a hydrophobic transmembrane spanning region, and a highly charged cytoplasmic tail, which is essential for RAGE ligand-mediated signal transduction (83, 87).

The extracellular fragment of RAGE consists of three immunoglobulin-like parts: a variable N-terminal V-type domain, followed by two constant C1- and C2-type domains. The V-type domain is essential for ligand binding. Interestingly, both in vitro and in vivo studies report that V-type and the adjacent C1-type domains are able to form structural and functional VC1 tandem unit with multiple binding sites to recognize numerous different RAGE ligands (87–89). These ligands interact with positively charged VC1 complex in a heterogeneous manner and, at first sight, they have no significant structural similarity. Nevertheless, a diverse range of RAGE ligands have at least one common feature – their acidic nature (negative charge) (90), though hydrophobicity-dependent mechanisms for RAGE-ligands interaction are also important (91).

In addition to the mentioned above, oligomerization of either RAGE or its ligands is one more key mechanism, controlling AGE-RAGE interplay (90). For example, oligomeric ligands can initiate self-association of the V-V/C1-C1regions in RAGE (Figure 4), forming relatively stable VC1-ligand complex with increased affinity to a diverse spectrum of ligands, including AGE-adducts and AGE-crosslinks. Also, such clustering shifts the equilibrium between monomer/oligomeric states of RAGE and respective ligands. This creates a positive feed-back loop and amplifies the RAGE ligand-mediated signal transduction. Under normal physiological conditions, RAGE expression is rather low. In contrast, during chronic inflammation RAGE is upregulated due to elevation of its ligand levels and the amplifying mechanisms of RAGE activation (92, 93). Generally, increased levels of AGEs, along with oligomerization and activation of RAGE, may explain why the RAGE signaling pathway is highly activated in various inflammation-related pathologies.

Obviously, the oligomerization is not the only mechanism underlying AGE-RAGE axis operation. Due to alternative splicing and metalloprotease-mediated cleavage, several variants of human RAGE appear, which regulate downstream pathways (84, 93). Figure 4 demonstrates the structures of the best-characterized RAGE isoforms: full-length RAGE (fl-RAGE), N-truncated RAGE (Nt-RAGE), dominant negative RAGE (DN-RAGE), and two soluble/secretory forms (sRAGE)—proteolytically cleaved RAGE (cRAGE) and C-truncated RAGE (endogenous secretory, esRAGE).

Deletion of the V-type domain transforms fl-RAGE into the Nt-RAGE variant, which is membrane-bound protein like fl-RAGE, but lacking the ability to bind AGEs. The Nt-RAGE protein is suggested to be involved in the control of angiogenesis by regulating AGE-independent migration of endothelial cells (94). Overexpression of Nt-RAGE may contribute to pathological events independently on the V-type domain.

The DN-RAGE form has the extracellular (V- and C-type regions) and transmembrane domains like fl-RAGE, but lacks the intracellular cytoplasmic domain required for initiating downstream signaling. Consequently, DN-RAGE binds AGE ligands, forming complexes that are not signaling-competent. Moreover, DN-RAGE can suppress RAGE-induced signaling by clustering with fl-RAGE and formation of hetero-complexes able to bind extracellular ligands without transducing signals into the cell. It is highly possible, that under certain conditions, DN-RAGE formation serves as an important mechanism to attenuate of fl-RAGE-induced signaling. For example, DN-RAGE has been found to inhibit tumor cell proliferation and invasion in vitro and tumorigenesis in vivo (95). Under normal physiological conditions, the level of DN-RAGE expression is similar to that of fl-RAGE (96).

Similar to DN-RAGE, sRAGE has no intracellular signaling tail, but also does not have the transmembrane domain, thus it is circulating protein. sRAGE is a heterogeneous group of proteins, resulting from either alternative splicing of fl-RAGE mRNA (esRAGE) or proteolytic cleavage of fl-RAGE from the cell surface (cRAGE) (84, 93). A unique 16-amino-acid segment in the C2 domain distinguishes esRAGE from cRAGE. The serum level of cRAGE is 5-fold higher than that of esRAGE in healthy individuals (97). Both sRAGE forms are believed to be competitive inhibitors of RAGE. By possessing the same AGE-binding ability as fl-RAGE, sRAGE acts as a “decoy” receptor, preventing circulating AGEs from interacting with RAGE (98), and thus providing cytoprotective effects against the AGE-RAGE complex negative impact. In addition, similar to DN-RAGE, esRAGE can interact with fl-RAGE, leading to the formation of hetero-complexes, which can bind extracellular ligands, but without transducing a signal into the cell.

Overall, understanding how various fructose-derived AGEs interact with multifaceted RAGE opens numerous ways to influence both the normal physiological and pathological events in the organism. Growing evidence supports a dual role of fructose itself, fructose-derived glycation products, and RAGE (21, 23, 93, 99–101). However, the AGE-RAGE axis is primarily known for its adverse effects on cellular metabolism and function through inflammation and nonenzymatic processes. For instance, the AGE-RAGE signaling pathway is believed to induce the expression of proinflammatory modulators (IL-6, TNF-α, IFN-γ) via activation of the transcription nuclear factor kappa-B (NF-κB) (15). Also, NF-κB increases RAGE expression, which in turn again stimulates NF-κB (88), creating a vicious cycle and generating more and more reactive species in new rounds (Figure 1). Additionally, RAGE expression induces NADPH oxidase-dependent generation of superoxide anion-radical and hydrogen peroxide (15), which contribute to the production of new AGEs as additional RAGE ligands. Overexpression of RAGE has been observed in various pathologies across a wide range of cells, including hepatocytes, cardiomyocytes, neuronal cells, vascular endothelial cells, and renal cells. Upregulation of RAGE was also found in such immune cells as neutrophils, monocytes/macrophages, lymphocytes, dendritic cells, where its activation resulted in the induction of expression of NF-κB-regulated genes encoding proinflammatory cytokines, chemokines, and adhesion molecules (15, 102–104).

At least the aforementioned points suggest that attenuation of the AGE-RAGE signaling could be a beneficial approach to preventing fructose-mediated disturbances. While fructose is a valuable, healthy, and safe nutrient at moderate intake levels (≤50 g/day or < 10% of total calories per day) (53–56), long-term consumption of excessive fructose and elevated endogenous its production have been shown to have detrimental impacts (19–23, 47, 48). There several potential targets within the pathway from fructose to active RAGE that could be considered for modulating the fructose-initiated AGE-RAGE axis (Figure 5, red arrows 1–7). These targets to some extension may serve as “stepping stones” in the following strategies: (1) prevention of RCS/AGE formation, (2) reduction of RCS/AGE levels, and (3) suppression of RAGE-mediated signaling.

Most importantly, the human organism itself possesses a highly developed and tightly regulated antiglycoxidation system that operates at various levels to protect against the adverse effects of AGE-RAGE axis overactivation. Also, it should be kept in mind that strategies to counteract the detrimental impacts of RAGE signaling should preserve beneficial effects of RAGE. In this regard, it makes sense to apply mild approaches to modulate the axis.

Over recent decades, the impact of fructose on human health has attracted growing attention from the scientific community. The “fructose hypothesis” suggests that fructose plays a causative role in various health problems. Indeed, compelling evidence indicates that long-term intake of excessive fructose can have deleterious side effects, which are believed to be attributed to glycation and fructose-derived AGEs. It is well established that AGEs are involved in upregulation of the specific receptor for AGEs (RAGE), amplifying RAGE-mediated signaling related to inflammation, metabolic disorders, chronic diseases, and aging. In contrast, short-term consumption of low/moderate doses of fructose are found to be beneficial, as it contributes to decrease of blood pressure, body mass index, the levels of triglycerides and HbA1c, while also improving glucose tolerance. Fructose is a valuable nutrient under certain pathophysiological conditions, particularly due to inducing mild carbonyl/oxidative stress, which in turn stimulates protective mechanisms.

Recent advances in studying the AGE-RAGE axis have significantly extended our understanding of its crucial role in human health. Nevertheless, several key issues remain to be addressed. Unfortunately, there are no standardized conventional methods to measure dietary AGEs and other glycotoxins, which limits inter-laboratory comparisons of the data obtained. This limitation hinders developing a unified dietary AGE/glycotoxin database. Standardized methods also could help determine whether there are differences between fructose-derived AGEs and those generated by other dietary carbohydrates, such as glucose or galactose. Understanding these distinctions may provide insights into the specific characteristics of AGE-RAGE interactions.

Additional intriguing questions arise regarding the beneficial effects of low doses of fructose. Specifically, does this impact involve AGE-RAGE signaling? If so, which RAGE isoforms are responsible for the benefits associated with fructose? Furthermore, is it possible to optimize the RAGE/sRAGE ratio by using appropriate doses of fructose and other natural compounds in our diets?

While the AGE-RAGE axis downregulation appears to be a promising approach for attenuating various disease conditions linked to RAGE-mediated inflammation, it is crucial to ensure that strategies aimed at counteracting RAGE signaling do not diminish its beneficial effects. Although the structure of the AGE-RAGE network is relatively well-studied, further research is needed to expand our understanding of the mechanisms underlying the operation of the AGE-RAGE axis and their relationship to human health.